Supplemental Materials for “The association between … Supplemental Materials for “The...

26
Supplemental Materials for “The association between hippocampal subfield volumes and education in cognitively normal older adults and amnestic mild cognitive impairment patients” by Kang et al. Supplemental Methods 1. Neuropsychological test 2. APO E genotyping 3. The definition of early and late mild cognitive impairment Supplemental Results 1. The stepwise regression analyses 2. Mediation analyses Supplemental Tables 1. Table S1. Non-amnestic cognitive functions of the study participants 2. Table S2. Clinical Characteristics of amnestic mild cognitive impairment subjects Supplemental Figures 1. Figure S1. Relationship between hippocampal subfield volumes and education (plots of simple regression) 2. Figure S2. Relationships between cognitive functions and hippocampal subfield volumes (plots of simple regression) 3. Figure S3. Relationships between cognitive functions and education (plots of simple regression) .

Transcript of Supplemental Materials for “The association between … Supplemental Materials for “The...

  • Supplemental Materials for “The association between hippocampal subfield

    volumes and education in cognitively normal older adults and amnestic mild

    cognitive impairment patients” by Kang et al.

    Supplemental Methods

    1. Neuropsychological test

    2. APO E genotyping

    3. The definition of early and late mild cognitive impairment

    Supplemental Results

    1. The stepwise regression analyses

    2. Mediation analyses

    Supplemental Tables

    1. Table S1. Non-amnestic cognitive functions of the study participants

    2. Table S2. Clinical Characteristics of amnestic mild cognitive impairment subjects

    Supplemental Figures

    1. Figure S1. Relationship between hippocampal subfield volumes and education

    (plots of simple regression)

    2. Figure S2. Relationships between cognitive functions and hippocampal subfield

    volumes (plots of simple regression)

    3. Figure S3. Relationships between cognitive functions and education (plots of simple

    regression)

    .

  • Supplemental Methods

    1. Neuropsychological evaluation

    Cognitive status was assessed by neuropsychological testing at the Seoul St. Mary’s

    Hospital, The Catholic University of Korea. The cognitive functions of all the subjects

    were assessed with the Korean version of Consortium to Establish a Registry for

    Alzheimer’s Disease (CERAD-K), including Verbal Fluency (VF), 15-item Boston Naming

    Test (BNT), MMSE-K, Word List Memory (WLM), Word List Recall (WLR), Word List

    Recognition (WLRc), Constructional Praxis (CP), and Constructional Recall (CR). The

    CERAD is the standardized clinical and neuropsychological assessment battery for the

    evaluation of patients with Alzheimer's disease. The results were reviewed by a

    neuropsychologist to determine whether there was evidence of cognitive impairment.

    The VF score is the number of animal names that the subject speaks for himself for one

    minute. The score range of BNT is from 0 to 15 points. The score range of MMSE-K is

    from 0 to 30 points. The score range of WLM is from 0 to 30 points. The score range of

    WLR is from 0 to 10 points. The score range of WLR is from 0 to 10 points. The score

    range of WLRc is from 0 to 10 points. The score range of CP is from 0 to 11 points. The

    score range of CR is from 0 to 11 points.

    2. APO E genotyping

    DNA was isolated from blood using the QIAmp Blood DNA Maxi Kit protocol (Qiagen,

    Valencia, CA). Genotypes for two APOE SNPs, rs429358 (E*4) and rs7412 (E*2) were

    determined using TaqMan SNP genotyping assays (Applied Biosystems, Foster City,

    California).

  • 3. The definition of early and late mild cognitive impairment

    (i) Late mild cognitive impairment defined by the report on memory impairment and

    performance on delayed verbal recall (CERAD-K Word List Recall) below 1.5 SD of the

    norm, (ii) Early mild cognitive impairment defined by the report on memory

    impairment and performance on delayed verbal recall (CERAD-K Word List Recall)

    between 1.0 SD and 1.5 SD below the norm.

    Supplemental Results

    1. The stepwise regression analyses

    1.1 Relationship between education and hippocampal subfield volumes 1.1.1 Lt. parasubiculum, both groups

    Result of Stepwise Backward Regression

    Analysis of Variance Table

    Model 1: left parasubiculum ~ age + education Model 2: left parasubiculum ~ age + gender + education + APOE genotype

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 65 11673.14

    2 63 11406.79 2 266.35 0.74 0.4833

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 117.44 72.69 162.19 22.41

    5.24 < 0.001

    age -0.72 -1.31 -0.12 0.30 -0.27 -0.5 -0.05 0.11 -2.41 0.019

    education -0.83 -1.44 -0.22 0.31 -0.31 -0.53 -0.08 0.11 -2.70 0.009

    R2=0.1601, adj.R2=0.1343, F=6.2, p=0.003, AIC =550.87 Res. Df, residual degrees of freedom; RSS, residual sum of squares; Df, degree of freedom; Sq, squares;

    Pr(>F), Probability of F value.

  • 1.1.2 Lt. presubiculum, both groups

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: left presubiculum ~ age + education Model 2: left presubiculum ~ age + gender + education + APOE genotype

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 65 132668.1

    2 63 131895.6 2 772.5 0.18 0.832

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 553.66 402.79 704.52 75.54

    7.33 < 0.001

    age -3.43 -5.43 -1.43 1.00 -0.38 -0.6 -0.16 0.11 -3.43 0.001

    education -2.47 -4.54 -0.40 1.04 -0.26 -0.48 -0.05 0.11 -2.38 0.02

    R2=0.2027, adj.R2=0.1782, F=8.26, p < 0.001, AIC =716.15

    1.1.3 Lt. subiculum, both groups

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: left subiculum ~ age + education Model 2: left subiculum ~ age + gender + education + APOE genotype

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 65 247721.8

    2 63 246508.9 2 1212.94 0.15 0.8567

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 868.76 662.61 1074.91 103.22

    8.42 < 0.001

    age -5.90 -8.64 -3.17 1.37 -0.46 -0.67 -0.25 0.11 -4.31 < 0.001

    education -3.28 -6.11 -0.45 1.42 -0.25 -0.46 -0.04 0.11 -2.32 0.024

    R2=0.2604, adj.R2=0.2376, F=11.44, p < 0.001, AIC =758.61

  • 1.1.4 Lt. parabiculum, aMCI group

    Analysis of Variance Table

    Model 1: left parasubiculum ~ age + gender + education + APOE genotype Model 2: left parasubiculum ~ age + gender + education + APOE genotype

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 30 4685.8

    2 30 4685.8 0 0

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 136.31 64.82 207.79 35.00

    3.89 0.001 age -0.99 -1.95 -0.02 0.47 -0.32 -0.63 -0.02 0.15 -2.09 0.045

    gender 8.63 -1.91 19.17 5.16 0.25 -0.04 0.55 0.15 1.67 0.105 education -1.21 -2.23 -0.19 0.50 -0.37 -0.67 -0.07 0.15 -2.43 0.021

    APOE genotype -7.48 -17.72 2.76 5.01 -0.23 -0.52 0.07 0.15 -1.49 0.146

    R2=0.3581, adj.R2=0.2725, F=4.18, p=0.008, AIC =282.72

    1.1.5 Lt. parabiculum, aMCI group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: left subiculum ~ age + education + APOE genotype Model 2: left subiculum ~ age + gender + education + APOE genotype

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 31 120000.3

    2 30 119200.9 1 799.46 0.2 0.657

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 994.71 639.35 1350.07 174.24

    5.71 < 0.001

    age -7.28 -12.03 -2.53 2.33 -0.46 -0.75 -0.17 0.15 -3.13 0.004

    education -5.22 -10.25 -0.20 2.47 -0.31 -0.59 -0.02 0.14 -2.12 0.042

    APOE genotype -51.14 -101.23 -1.05 24.56 -0.3 -0.58 -0.02 0.14 -2.08 0.046

    R2=0.3933, adj.R2=0.3346, F=6.7, p=0.001, AIC =394.22

  • 1.2 Relationship between cognitive functions and hippocampal subfield volumes

    1.2.1 CERAD-K Word List Recall, Control Group

    1.2.1.1 Lt. parasubiculum, Control group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: left parasubiculum ~ age + APOE genotype + WLR Model 2: left parasubiculum ~ age + gender + education + APOE genotype + WLR

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 29 4219.51

    2 27 4100.11 2 119.4 0.39 0.6787

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 177.51 96.80 258.22 39.46

    4.50 < 0.001

    age -1.16 -2.10 -0.22 0.46 -0.46 -0.82 -0.1 0.18 -2.51 0.018

    APOE genotype -9.57 -20.97 1.82 5.57 -0.27 -0.57 0.04 0.16 -1.72 0.096

    WLR -6.13 -9.68 -2.57 1.74 -0.65 -1.01 -0.29 0.18 -3.53 0.001

    R2=0.3242, adj.R2=0.2543, F=4.64, p=0.009, AIC =263.73

  • 1.2.1.2 Lt. presubiculum, Control group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: left presubiculum ~ age + WLR Model 2: left presubiculum ~ age + gender + education + APOE genotype + WLR

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 30 51168.26

    2 27 49405.91 3 1762.35 0.32 0.8101

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 682.83 414.19 951.46 131.54

    5.19 < 0.001 age -3.85 -7.01 -0.69 1.55 -0.45 -0.8 -0.1 0.18 -2.49 0.018

    WLR -20.89 -32.82 -8.97 5.84 -0.64 -1 -0.29 0.18 -3.58 0.001

    R2=0.3065, adj.R2=0.2602, F=6.63, p=0.004, AIC =344.08

    1.2.1.3 Lt. subiculum, Control group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: left subiculum ~ age + WLR Model 2: left subiculum ~ age + gender + education + APOE genotype + WLR

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 30 79895.44

    2 27 70849.75 3 9045.7 1.15 0.3473

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 858.80 523.12 1194.48 164.36

    5.22 < 0.001 age -4.82 -8.77 -0.88 1.93 -0.49 -0.87 -0.1 0.19 -2.50 0.018

    WLR -15.72 -30.62 -0.81 7.30 -0.42 -0.8 -0.04 0.19 -2.15 0.039

    R2=0.1927, adj.R2=0.1389, F=3.58, p=0.04, AIC =358.78

  • 1.2.1.4 Rt. presubiculum, Control group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: right presubiculum ~ age + WLR Model 2: right presubiculum ~ age + gender + education + APOE genotype + WLR

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 30 51834.81

    2 27 48226.75 3 3608.07 0.67 0.5759

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 583.75 313.37 854.13 132.39

    4.41 < 0.001

    age -3.28 -6.45 -0.10 1.56 -0.42 -0.8 -0.03 0.2 -2.11 0.044

    WLR -12.76 -24.76 -0.75 5.88 -0.43 -0.82 -0.04 0.2 -2.17 0.038

    R2=0.1656, adj.R2=0.11, F=2.98, p=0.066, AIC =344.51

    1.2.1.5 Rt. subiculum, Control group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: right subiculum ~ age + WLR Model 2: right subiculum ~ age + gender + education + APOE genotype + WLR

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 30 66354.60

    2 27 61134.16 3 5220.44 0.77 0.5217

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 898.75 592.84 1204.66 149.79

    6.00 < 0.001 age -5.31 -8.90 -1.71 1.76 -0.57 -0.94 -0.2 0.19 -3.02 0.005

    WLR -14.33 -27.92 -0.75 6.65 -0.41 -0.78 -0.04 0.19 -2.16 0.039

    R2=0.2405, adj.R2=0.1899, F=4.75, p=0.016, AIC =352.66

  • 1.2.2 CERAD-K Word List Recognition, Control Group

    1.2.2.1 Lt. CA1, Control group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: left CA1 ~ gender + education + WLRc Model 2: left CA1 ~ age + gender + education + APOE genotype + WLRc

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 29 95537.57

    2 27 92372.32 2 3165.25 0.46 0.6345

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 290.25 124.83 455.66 80.88

    3.59 0.001

    gender 40.63 -7.56 88.81 23.56 0.3 -0.04 0.64 0.17 1.72 0.095

    education 2.90 -0.98 6.78 1.90 0.26 -0.07 0.59 0.17 1.53 0.137 WLRc 28.73 10.37 47.09 8.98 0.48 0.19 0.78 0.15 3.20 0.003

    R2=0.3644, adj.R2=0.2987, F=5.54, p=0.004, AIC =366.69

    1.2.2.2 Lt. CA3, Control group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: left CA3 ~ WLRc Model 2: left CA3 ~ age + gender + education + APOE genotype + WLRc

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 31 14005.46

    2 27 13052.57 4 952.89 0.49 0.741

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 88.51 28.95 148.08 29.21

    3.03 0.005

    WLRc 12.47 5.84 19.10 3.25 0.57 0.28 0.86 0.15 3.84 0.001

    R2=0.3218, adj.R2=0.2999, F=14.71, p < 0.001, AIC =299.32

  • 1.2.2.3 Lt. CA4, Control group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: left CA4 ~ WLRc Model 2: left CA4 ~ age + gender + education + APOE genotype + WLRc

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 31 23578.30

    2 27 21196.76 4 2381.54 0.76 0.561

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 115.72 38.43 193.00 37.89

    3.05 0.005

    WLRc 15.21 6.60 23.81 4.22 0.54 0.25 0.84 0.15 3.60 0.001

    R2=0.2954, adj.R2=0.2726, F=12.99, p=0.001, AIC =316.51

    1.2.2.4 Lt. GC-DG, Control group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: left GC-DG ~ APOE genotype + WLRc Model 2: left GC-DG ~ age + gender + education + APOE genotype + WLRc

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 30 34513.86

    2 27 31987.37 3 2526.49 0.71 0.554

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 133.70 37.31 230.09 47.20

    2.83 0.008

    APOE genotype 22.50 -8.92 53.93 15.39 0.23 -0.08 0.53 0.16 1.46 0.154

    WLRc 16.83 6.18 27.48 5.22 0.5 0.2 0.81 0.16 3.23 0.003

    R2=0.2809, adj.R2=0.233, F=5.86, p=0.007, AIC =331.09

  • 1.2.2.5 Lt. subiculum, Control group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: left subiculum ~ APOE genotype + WLRc Model 2: left subiculum ~ age + gender + education + APOE_genotype1 + WLRc

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 30 78738.69

    2 27 73099.10 3 5639.58 0.69 0.5635

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 272.12 126.53 417.71 71.29

    3.82 0.001

    APOE genotype 47.85 0.39 95.31 23.24 0.34 0.02 0.66 0.16 2.06 0.048

    WLRc 16.22 0.14 32.31 7.88 0.34 0.02 0.66 0.16 2.06 0.048

    R2=0.2044, adj.R2=0.1513, F=3.85, p=0.032, AIC =358.3 1.2.2.6 Lt. total hippocampus, Control group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: left total hippocampus ~ WLRc Model 2: left total hippocampus ~ age + gender + education + APOE genotype + WLRc

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 31 3247719

    2 27 2828408 4 419310.99 1 0.4244

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 1882.69 975.64 2789.73 444.74

    4.23 < 0.001

    WLRc 150.21 49.21 251.20 49.52 0.48 0.17 0.79 0.16 3.03 0.005

    R2=0.2289, adj.R2=0.204, F=9.2, p=0.005, AIC =479.05

  • 1.2.2.7 Rt. CA4, Control group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: right CA4 ~ gender + education + WLRc Model 2: right CA4 ~ age + gender + education + APOE genotype + WLRc

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 29 19645.39

    2 27 19127.14 2 518.24 0.37 0.697

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 161.46 86.45 236.47 36.67

    4.40 < 0.001

    gender 19.70 -2.15 41.55 10.68 0.34 -0.02 0.7 0.18 1.84 0.075

    education 1.53 -0.22 3.29 0.86 0.32 -0.03 0.68 0.18 1.78 0.085

    WLRc 9.18 0.85 17.51 4.07 0.36 0.05 0.68 0.16 2.26 0.032

    R2=0.2841, adj.R2=0.21, F=3.84, p=0.02, AIC =314.49

    1.2.2.8 Rt. total hippocampus, Control group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: right total hippocampus ~ age + WLRc Model 2: right total hippocampus ~ age + gender + education + APOE genotype + WLRc

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 30 2903657

    2 27 2590957 3 312699.66 1.09 0.3717

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 3752.70 1837.95 5667.45 937.56

    4.00 < 0.001

    age -19.10 -39.77 1.57 10.12 -0.3 -0.61 0.01 0.16 -1.89 0.069

    WLRc 113.01 12.86 213.17 49.04 0.37 0.05 0.68 0.16 2.30 0.028

    R2=0.2796, adj.R2=0.2316, F=5.82, p=0.007, AIC =477.35

  • 1.2.3 CERAD-K Constructional Recall, aMCI Group

    1.2.3.1 Lt. GC-DG, aMCI group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: left GC-DG ~ education + CR Model 2: left GC-DG ~ age + gender + education + APOE genotype + CR

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 32 49308.70

    2 29 45777.43 3 3531.28 0.75 0.5336

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 270.29 234.75 305.82 17.44

    15.49 < 0.001

    education -3.75 -6.87 -0.63 1.53 -0.38 -0.68 -0.08 0.15 -2.45 0.02

    CR 7.55 1.96 13.14 2.74 0.43 0.12 0.73 0.15 2.75 0.01

    R2=0.262, adj.R2=0.2159, F=5.68, p=0.008, AIC =361.09

    1.2.3.2 Rt. CA4, aMCI group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: right CA4 ~ education + CR Model 2: right CA4 ~ age + gender + education + APOE genotype + CR

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 32 37898.05

    2 29 35757.42 3 2140.64 0.58 0.6337

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 257.78 226.63 288.93 15.29

    16.86 < 0.001

    education -2.50 -5.24 0.24 1.34 -0.3 -0.61 0.02 0.16 -1.86 0.072

    CR 6.24 1.34 11.14 2.41 0.41 0.1 0.73 0.16 2.59 0.014

    R2=0.2123, adj.R2=0.1631, F=4.31, p=0.022, AIC =351.88

  • 1.2.3.3 Rt. GC-DG, aMCI group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: right GC-DG ~ gender + education + CR Model 2: right GC-DG ~ age + gender + education + APOE genotype + CR

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 31 54838.15

    2 29 52643.64 2 2194.51 0.6 0.5531

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 314.98 262.91 367.04 25.53

    12.34 < 0.001 gender -24.15 -60.37 12.08 17.76 -0.22 -0.53 0.1 0.16 -1.36 0.184

    education -3.41 -6.76 -0.06 1.64 -0.32 -0.62 -0.02 0.15 -2.07 0.046 CR 7.59 1.32 13.85 3.07 0.4 0.08 0.71 0.16 2.47 0.019

    R2=0.295, adj.R2=0.2267, F=4.32, p=0.012, AIC =366.81

    1.2.3.4 Rt. presubiculum, aMCI group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: right presubiculum ~ education + CR Model 2: right presubiculum ~ age + gender + education + APOE genotype + CR

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 32 91858.77

    2 29 85751.19 3 6107.58 0.69 0.5664

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std.Beta lwr upr SE t value p

    (Intercept) 270.36 221.86 318.85 23.81

    11.36 < 0.001 education -5.18 -9.44 -0.92 2.09 -0.38 -0.68 -0.08 0.15 -2.48 0.019

    CR 11.16 3.53 18.79 3.75 0.45 0.16 0.75 0.15 2.98 0.005

    R2=0.2825, adj.R2=0.2377, F=6.3, p=0.005, AIC =382.87

  • 1.2.3.5 Rt. total hippocampus, aMCI group

    Result of Stepwise Backward Regression

    Analysis of Variance Table Model 1: right total hippocampus ~ education + CR Model 2: right total hippocampus ~ age + gender + education + APOE genotype + CR

    Res.Df RSS Df Sum of Sq F Pr(>F)

    1 32 8509529

    2 29 7828342 3 681186.97 0.84 0.4825

    Finally Selected Model

    Result of Multiple Regression Analysis

    Beta lwr upr SE std. Beta lwr upr SE t value p

    (Intercept) 3295.09 2828.32 3761.86 229.16

    14.38 < 0.001

    education -41.72 -82.73 -0.72 20.13 -0.33 -0.64 -0.02 0.16 -2.07 0.046

    CR 93.67 20.24 167.10 36.05 0.41 0.1 0.72 0.16 2.60 0.014

    R2=0.2252, adj.R2=0.1767, F=4.65, p=0.017, AIC =541.37

    2. Mediation Analyses

    2.1 The proposed mediator: Left GC-DG

    Indirect effect of education on CERAD-K CR score

    Effect Boot SE BootLLCI BootULCI

    Left GC-DG .0047 .0369 -.0547 .1008

    Boot, bootstrapping; SE, standard error; LLCI, lower limit of confidence interval; ULCI, upper limit of

    confidence interval.

    Normal Theory tests for indirect effect

    Effect SE Z p

    Left GC-DG .0047 .0358 .1320 .8950

    SE, standard error.

  • 2.2 The proposed mediator: Right CA4

    Indirect effect of education on CERAD-K CR score

    Effect Boot SE BootLLCI BootULCI

    Right CA4 .0154 .0298 -.0337 .0880

    Boot, bootstrapping; SE, standard error; LLCI, lower limit of confidence interval; ULCI, upper limit of

    confidence interval.

    Normal Theory tests for indirect effect

    Effect SE Z p

    Right CA4 .0154 .0311 .4956 .6202

    SE, standard error.

    2.3 The proposed mediator: Right GC-DG

    Indirect effect of education on CERAD-K CR score

    Effect Boot SE BootLLCI BootULCI

    Right GC-DG .0116 .0313 -.0414 .0895

    Boot, bootstrapping; SE, standard error; LLCI, lower limit of confidence interval; ULCI, upper limit of

    confidence interval.

    Normal Theory tests for indirect effect

    Effect SE Z p

    Right GC-DG .0116 .0308 .3761 .7069

    SE, standard error.

  • 2.4 The proposed mediator: Right presubiculum

    Indirect effect of education on CERAD-K CR score

    Effect Boot SE BootLLCI BootULCI

    Right presubiculum -.0086 .0309 -.0915 .0433

    Boot, bootstrapping; SE, standard error; LLCI, lower limit of confidence interval; ULCI, upper limit of

    confidence interval.

    Normal Theory tests for indirect effect

    Effect SE Z p

    Right presubiculum -.00086 .0308 -.2807 .7790

    SE, standard error.

    2.4 The proposed mediator: Right total hippocampus

    Indirect effect of education on CERAD-K CR score

    Effect Boot SE BootLLCI BootULCI

    Right total hippocampus -.0045 .0261 -.0351 .0732

    Boot, bootstrapping; SE, standard error; LLCI, lower limit of confidence interval; ULCI, upper limit of

    confidence interval.

    Normal Theory tests for indirect effect

    Effect SE Z p

    Right total hippocampus .0045 .0273 .1643 .8695

    SE, standard error.

  • Supplemental Tables

    Table S1. Non-amnestic cognitive functions of the study participants

    Control Group

    (n=39)

    aMCI group

    (n=38)

    P value

    CERAD-K VF 14.3 ± 3.7 (8-24) 12.5 ± 3.4 (7-21) 0.033*

    CERAD-K BNT 11.8 ± 2.7 (5-15) 10.8 ± 2.4 (5-15) 0.093

    CERAD-K CP 9.9 ± 1.5 (6-11) 9.7 ± 1.6 (5-11) 0.461

    CERAD-K, the Korean version of Consortium to Establish a Registry for Alzheimer’s Disease; VF, Verbal

    Fluency; BNT, 15-item Boston Naming Test; CP, Constructional Praxis. *p < 0.05

    Table S2. Clinical Characteristics of amnestic mild cognitive impairment subjects

    aMCI group

    (n=38)

    Early/late aMCI (n, %) 21 : 17 (55.3 : 44.7)

    aMCI, Single/Multiple domain (n, %)

    4 : 34 (10.5 : 89.5)

    CERAD-K WLM Impairment (n, %)

    23 (60.5)

    CERAD-K WLR Impairment (n, %)

    34 (89.5)

    CERAD-K WLRc Impairment (n, %)

    24 (63.2)

    CERAD-K CR Impairment (n, %)

    22 (57.9)

    CERAD-K, the Korean version of Consortium to Establish a Registry for Alzheimer’s Disease; WLM, Word

    List Memory; WLR, Word List Recall; WLRc, Word List Recognition; CR, Constructional Recall.

  • Supplemental Figures

    Figure S1. Relationship between hippocampal subfield volumes and education (plots of simple regression)

    (A) Relationship between education and Lt. parasubiculum in both groups

    (B) Relationship between education and Lt. presubiculum in both groups

  • (C) Relationship between education and Lt. subiculum in both groups

    (D) Relationship between education and Lt. parasubiculum in aMCI group

  • (E) Relationship between education and Lt. subiculum in aMCI group

  • Figure S2. Relationships between cognitive functions and hippocampal subfield volumes (plots of simple regression)

    (A) Relationship between CERAD-K WLR and Lt. parasubiculum in control group

    (B) Relationship between CERAD-K WLR and Lt. presubiculum in control group

  • (C) Relationship between CERAD-K WLRc and Lt. CA3 in control group

    (D) Relationship between CERAD-K WLRc and Lt. CA4 in control group

  • (E) Relationship between CERAD-K CR and Rt. presubiculum in aMCI group

  • Figure S3. Relationships between cognitive functions and education (plots of simple regression)

    (A) Relationship between education and MMSE-K in control group

    (B) Relationship between education and CERAD-K CR in control group

  • (C) Relationship between education and MMSE-K in aMCI group

    (D) Relationship between education and CERAD-K CR in aMCI group

    CERAD-K, the Korean version of Consortium to Establish a Registry for Alzheimer’s Disease; MMSE-K, The Korean version of mini mental status examination; WLR, Word List Recall; WLRc, Word List Recognition; CR, Constructional Recall